Snipers, Shills, and Sharks: eBay and Human Behavior
Ken Steiglitz, Professor of Computer Science at Princeton University and an avid collector of ancient coins, has combined his interests in a new book Snipers, Shills, and Sharks: eBay and Human Behavior. “Why is eBay so successful?” and “How well does the theory of auctions work in predicting human behavior?” were two of the questions he addressed at the October 17 Lunch ‘n Learn seminar.
eBay has succeeded, Steiglitz suggests, not simply because it began early but more importantly because its operation engenders trust. Feedback mechanisms provide a degree of trust to both buyers and sellers, and eBay offers transparency to all involved. Only recently have they retreated somewhat by making bids over $200 anonymous and by eliminating searches by bidder to active auctions (a change that has just been reversed).
Steiglitz reviewed the theoretical roots of auction theory, notably papers by William Vickrey in 1961 (he won the Nobel prize in 1999), and by Riley and Samuelson (1981), Myerson in 1981 (and he won the Nobel prize just last week), and Milgrom and Weber (1982). Other Nobel prizes are associated with auction theory, behavorial economics, and mechanism design in general: Vernon Smith, Daniel Kahneman, John Nash. In fact, this general field has been among the most successful and active parts of economics for several decades.
Steiglitz notes that beginners to eBay tend to bid early, taking advantage of eBay’s proxy bidding. By so doing, he emphasizes, others are likely to bid more often to test the limits of the original bid. Far more strategic is the practice of sniping, essentially placing the bid only within the final seconds of the auction.
Here is the danger of early bidding. Bidder #1 has placed a very early bid. Bidder #2, in two separate sessions, places many bids that never succeed in achieving the highest bid. We do not know how high bidder #1’s actual bid was. In eBay auctions, the winner will obtain the item for an increment above the second highest bid. By placing the bid early, bidder 1 may have paid much more for the item than had the bid been placed at the very end of the auction.
He emphasizes that buyers ought to snipe in order to decrease competition, essentially to lay low and not to reveal their expertise or interest. By contrast, sellers should act to encourage competition, by using low opening bids, avoiding secret reserves, and of course by being informative and nice to bidders and buyers. It takes courage, of course, to offer expensive items at a low opening bid, but the free market often offers comfort.
The internet offers a goldmine of data for empirical studies that help to augment and extend auction theory. One such study by Katkar & Lucking-Reiley (2000) demonstrated that bidders tend to avoid items with secret reserves. Others suggest that buyers tend somewhat to overlook higher shipping costs and that bidders tend to prefer auctions with many bidders (herding behavior).
Steiglitz concludes with a few tales from the dark side of online auctions. There are shills who enter bids that aim to protect the value of the dealer’s items. There are forgeries and fakes. And perhaps the hardest to prove, there may be rings with colluding bidders.
Steiglitz concludes that human behavior can be much more complex than that predicted by theory. Nonetheless, he continues to pursue his collecting passion, drawing upon his expertise about both auctions and the coins he collects.
The presentation and a podcast are available.
Please feel free to use the comments feature to share your own eBay war stories!